Skip to main content
ABC News
The Top Seeds To Trust — And Avoid — In The Men’s NCAA Tournament

From the outside, March 16, 2012, was the maddest day in the history of the men’s NCAA Tournament. Before that tourney, only four No. 2 seeds had ever lost their opening-round tilts against No. 15 seeds. On that day, it happened not once, but twice.

The first Goliath to fall was Missouri, a historically solid program that had experienced a resurgence in its last season in the Big 12 Conference. Anchored by a hot-shooting, four-guard lineup, the Tigers entered their game against Norfolk State with just a 2.9 percent chance of losing, according to Ken Pomeroy. Norfolk State’s Kyle O’Quinn made that outside shot a reality, posting a dominant 26-point, 14-rebound performance en route to the Spartans’ shocker.

What makes LeBron James so great | FiveThirtyEight

The second and arguably more headline-grabbing upset was that of Duke, whose tournament resume needs no introduction. The Blue Devils had a potent 3-point attack led by future NBA players Austin Rivers and Seth Curry, yet they couldn’t stop the Lehigh Mountain Hawks’ future pro — CJ McCollum, now starring for the Portland Trail Blazers — on the other end. The junior torched Mike Krzyzewski’s squad to the tune of 30 points, six rebounds and six assists, upending the then-four-time national champions in their home state.

In retrospect, though, there were clear red flags for each favorite heading into the tournament. The Tigers’ No. 1 offense obscured a defense that ranked 80th in adjusted defensive efficiency, and the Blue Devils didn’t fare much better on that end, posting the 71st-best defense. While a first-round exit seemed unlikely even to the programs’ biggest detractors,1 a relatively early departure from the sport’s biggest stage was strongly suggested by prior history.

An illustration of a basketball where the lines are the brackets

related: 2021 Men’s and Women’s March Madness Predictions Read more. »

Since KenPom started tracking adjusted efficiency stats in the 2001-02 season, only four out of 72 Final Four teams have posted a regular-season defensive efficiency ranking worse than 50. There’s been a little more leeway on offense — eight of the 72 Final Four teams have posted offensive efficiency rankings outside the top 50 since 2002. But only one title winner has finished the regular season with a ranking on either metric outside the top 50: Connecticut in 2014 had the 12th best defensive efficiency rating but was just 57th on offense.

Given how essential strong efficiency ratings seem to be, which highly seeded teams this year have the potential to go far — and which shouldn’t count on cutting down the nets? We looked back over the past 18 tournaments2 and categorized the teams at the top four seed lines into those teams that met expectations — for example, a No. 1 seed making the Final Four, a No. 2 seed making the Elite Eight, etc.3 — and those that fell short. From there, we found the average pre-tournament adjusted efficiency rankings for each group, and then determined where this year’s top seeds fell in those groups.

Which teams are candidates for an upset?

Top four seed lines in the 2021 men’s NCAA Tournament by their rankings in adjusted offensive and defensive efficiency and the difference between those rankings and the average underperforming team at that seed

vs. underperformer
Team笆イ笆シ seed笆イ笆シ Adj. off. rk笆イ笆シ Adj. def. rk笆イ笆シ Off.笆イ笆シ Def.笆イ笆シ
Ohio St. 2 4 79 +12.1 -52.4
Baylor 1 3 44 +9.1 -31.7
West Virginia 3 11 65 +18.6 -31.3
Iowa 2 2 50 +14.1 -23.4
Florida St. 4 10 48 +18.8 -11.7
Texas 3 21 36 +8.6 -2.3
Gonzaga 1 1 10 +11.1 +2.3
Virginia 4 12 33 +16.8 +3.3
Michigan 1 6 7 +6.1 +5.3
Illinois 1 7 5 +5.1 +7.3
Houston 2 8 16 +8.1 +10.6
Purdue 4 23 23 +5.8 +13.3
Oklahoma St. 4 54 22 -25.2 +14.3
Arkansas 3 35 14 -5.4 +19.7
Alabama 2 34 2 -17.9 +24.6
Kansas 3 59 6 -29.4 +27.7

Underperformers are teams that lost in the tournament before their seed line would suggest they should. Average rankings taken of seeds since the 2002 tournament. Efficiency rankings are as of March 17.


History isn’t guaranteed to repeat itself — especially in a one-of-a-kind tournament in 2021. But the last couple decades can offer some useful hints about which teams to trust and which to be suspicious of in the Hoosier State.

1 seeds

vs. underperformer
Team Adj. off. rk Adj. def. rk Offense Defense
Baylor 3 44 +9.1 -31.7
Gonzaga 1 10 +11.1 +2.3
Michigan 6 7 +6.1 +5.3
Illinois 7 5 +5.1 +7.3


Since 2002, 45 No. 1 seeds underperformed their seed expectation by losing before the Final Four, and those 45 had an average KenPom offensive efficiency ranking of 12.1 and a defensive efficiency ranking of 12.3. Of the 27 No. 1 seeds that met expectations, the average offensive efficiency rating has been 6.0, while the typical defensive ranking has been 13.2.4

Among this year’s 1 seeds, undefeated Gonzaga (No. 1 offense, No. 10 defense), Michigan (No. 6 offense, No. 7 defense) and Illinois (No. 7 offense, No. 5 defense) all fall within or near the “meeting expectation” boundaries of the last two decades.

A promotional collage of basketball players in black and white with only orange on their ball/jerseys colored in.

related: The Favorites And Cinderellas Of The Men’s NCAA Tournament Read more. »

This year’s Zags sport Mark Few’s best offensive unit, and they are also a disruptive bunch — behind freshman guard Jalen Suggs’s two steals per game, the Bulldogs have the 42nd-best steal rate in the country. The Wolverines have been reeling recently, dropping three of their past five games and losing senior forward Isaiah Livers to a foot injury in the Big Ten Tournament. But even though Michigan boasts the worst title odds (3 percent) of any No. 1 seed in FiveThirtyEight’s men’s March Madness predictions, it has the rare balance of elite offense and defense that’s come to define deep runs in the Big Dance. Illinois has effectively used what may be the best inside-out duo in the nation in Kofi Cockburn and Ayo Dosunmu, capturing the conference tournament title in the loaded Big Ten.

Baylor (No. 3 offense, No. 44 defense), however, is in riskier territory. It has struggled on the defensive end since returning from a COVID-19 pause, surrendering at least 1 point per possession in its last seven games. For a team that forged its identity for much of the past two seasons around its defense, it may come as a surprise that the Bears’ Achilles’ heel would be on that end of the court. The Bears have hope yet, though: Two of the five No. 1 seeds with Baylor’s defensive ranking or worse have made the Final Four, though neither won the title. The lowest-ranked defense to win a title was 37th — a tie between 2015 Duke and 2009 North Carolina, which had this mark entering their respective NCAA Tournaments.

2 seeds

vs. underperformer
Team Adj. off. rk Adj. def. rk Offense Defense
Ohio St. 4 79 +12.1 -52.4
Iowa 2 50 +14.1 -23.4
Houston 8 16 +8.1 +10.6
Alabama 34 2 -17.9 +24.6


The typical underperforming 2 seed has had an average offensive efficiency ranking of 16.1 and a defensive ranking of 26.6, while the average 2 seed to at least make the Elite Eight had an average of 15.0 and 17.9, respectively.

By those measures, this year’s crop of 2 seeds has some potential underachievers. Iowa (No. 2 offense, No. 50 defense), Ohio State (No. 4 offense, No. 79 defense) and Alabama (No. 34 offense, No. 2 defense) each have at least one unit that leaves something to be desired. While the Hawkeyes are anchored down low by National Player of the Year favorite Luka Garza, their porous defense has often kept lesser opponents in games. The Buckeyes, meanwhile, are even more sieve-like on that end; the last team to make the Final Four with a defense as bad as the Buckeyes’ was the havoc machine of Virginia Commonwealth in 2011. The Crimson Tide have the opposite quandary, as they lock down opposing shooters, giving up the 10th-lowest effective field-goal percentage, but struggle mightily to score the basketball.

Every baseball fan should know these Negro League stars | FiveThirtyEight

The one No. 2 seed that looks well-positioned to make a deep run based on its numbers is the Houston Cougars (No. 8 offense, No. 16 defense). With a stingy defense and a steady offense conducted by sharpshooter Quentin Grimes, the Cougars haven’t looked this good since the days of Phi Slama Jama. And critically, two areas in which the Cougars excel — offensive rebounding and not turning the ball over — are characteristic of NCAA Tournament favorites that avoid upsets.

3 seeds 

vs. underperformer
Team Adj. off. rk Adj. def. rk Offense Defense
West Virginia 11 65 +18.6 -31.3
Texas 21 36 +8.6 -2.3
Arkansas 35 14 -5.4 +19.7
Kansas 59 6 -29.4 +27.7


The average underperforming No. 3 seed has placed 29.6 and 33.7 in Pomeroy’s offensive and defensive rankings, while the average 3 seed to make it at least as far as the Sweet 16 has ranked 18.8 and 31.4. 

Once again, there is an array of offensive and defensive strengths among a seed line. Big 12 foes Kansas (No. 59 offense, No. 6 defense) and West Virginia (No. 11 offense, No. 65 defense) have prioritized opposite ends of the floor, with the Jayhawks holding teams under 46 percent on 2-pointers and the Mountaineers remaking their identity around superior spacing and outside shooting than in years prior. As it stands, just three No. 3 seeds in the KenPom era have had an offensive efficiency ranking as bad as this Kansas team’s, and none of them made it past the Round of 32. History is more on the side of the Mountaineers: Out of nine No. 3 seeds since 2002 with West Virginia’s defensive ranking or worse, five have made at least the Sweet 16.

A collage of three college women’s basketball players against shapes, all greyed out except for the orange basketball they’re holding.

related: The Top Picks And Sleepers To Watch In The Women’s NCAA Tournament Read more. »

Arkansas (No. 35 offense, No. 14 defense) is slightly more balanced than either of the aforementioned teams — though it faces some of the same offensive concerns as Kansas — and has its highest NCAA Tournament seed since 1995 under second-year coach Eric Musselman. Texas (No. 21 offense, No. 36 defense) posted middling metrics this season relative to past No. 3 seeds but has surged as of late, winning its last five and capturing the Big 12 title.

4 seeds

vs. underperformer
Team Adj. off. rk Adj. def. rk Offense Defense
Florida St. 10 48 +18.8 -11.7
Virginia 12 33 +16.8 +3.3
Purdue 23 23 +5.8 +13.3
Oklahoma St. 54 22 -25.2 +14.3


The average No. 4 seed that failed to make at least the Sweet 16 had an offensive efficiency ranking of 28.8 and a defensive ranking of 36.3, while its Sweet 16 counterparts had efficiency rankings of 28.2 and 28.7.

Though there isn’t much of a difference on offense between No. 4 seeds that met expectations and those that flamed out, there is a slight difference on defense — and this crop of No. 4 seeds has a couple stragglers on both sides of the ball. That doesn’t include Virginia (No. 12 offense, No. 33 defense), which is much weaker on defense than we’ve come to expect from Tony Bennett-coached teams but whose metrics bode well for a chance for a rematch with No. 1 Gonzaga in the Sweet 16 (COVID-19 issues notwithstanding). Neither does it include Purdue (No. 23 offense and defense), which is even-steven on both fronts and might have a secret advantage not captured by KenPom: playing in their home state of Indiana.

The No. 4 seeds that might run into some trouble are Florida State (No. 10 offense, No. 48 defense) and Oklahoma State (No. 54 offense, No. 22 defense). The Seminoles have the tallest team in the country, according to KenPom’s effective height metric, which leads to the eighth-best block percentage in the country. But they are incredibly foul-prone, ranking in the bottom 50 of defensive free-throw rate, and don’t box out all that well (opponents grab 31.4 percent of their misses). The Cowboys have defended well enough against a tough Big 12 Conference slate, but they struggle mightily on offense despite the brilliance of projected No. 1 overall pick Cade Cunningham. Oklahoma State turns the ball over on a staggering 21.5 percent of its possessions, and when you combine that with subpar defensive rebounding, it’s easy to envision the Cowboys prematurely exiting stage left.

If you’re just looking at who has the best shot at making a deep run in the tournament, regardless of seeding expectation, the KenPom end-of-season top 25 should suffice. Since KenPom starting tracking adjusted efficiency stats in the 2001-02 season, every single men’s basketball champion has featured a pre-tournament ranking of at least 25, and only three — the Carmelo Anthony-led Syracuse Orange in 2003 and the Kemba Walker– and Shabazz Napier-led Connecticut Huskies of 2011 and 2014, respectively — were outside the top 10. If you broaden that out to all Final Four teams in that time period, just six of 72 ranked outside of the top 25 before the Big Dance. That could spell trouble for the Big 12 trio of West Virginia, Oklahoma State and Texas, who are all outside the top 25 as the tourney begins.

Within that pre-tournament top 25, the teams that bring home hardware tend to be balanced between their offensive and defensive capabilities. Thirteen of the past 18 champions have ranked in the top 25 of both adjusted offense and defense before the tournament. This year, top seeds Gonzaga, Michigan, Illinois, Houston and Purdue fit the profile of recent teams who’ve made deep runs into the tournament. 

To put it differently, the tournament isn’t truly anyone’s game, and almost all eventual title winners had the regular-season numbers to suggest they were worthy contenders. If one of the top teams has an especially porous defense, it just might find itself on the losing end of an upset for the ages, adding to the chaos of March.

Check out our latest March Madness predictions.

Who’s the most chaotic fictional football coach? | FiveThirtyEight

What makes Patrick Mahomes so great? | FiveThirtyEight


  1. Looking at you, Kansas and North Carolina fans.

  2. From 2002 to 2019. The 2020 NCAA Tournament was canceled due to the COVID-19 pandemic.

  3. A No. 1 seed that gets to the Final Four and loses to a No. 3 seed still meets its seeding expectations, since it will have gotten past each stage of the tournament in which it was guaranteed to face a lower-seeded team.

  4. The lone No. 1 seed to lose to a No. 16 seed, the 2018 Virginia Cavaliers, had the 21st-best offense and top-rated defense entering the tournament.

Santul Nerkar was a copy editor at FiveThirtyEight.